Hasil untuk "Specialties of internal medicine"

Sylvain Diop, Elie Fadel, Thibaut Genty et al.

Objectives: Cardiac surgery is associated with an increased dioxygen (O2) consumption (VO2) following cardiopulmonary bypass (CPB). But data on intraoperative VO2 variation during pulmonary endarterectomy (PEA) are scarce. We aimed to assess the variation of VO2 and O2 delivery (DO2) between the induction of general anesthesia and the weaning off CPB in patients undergoing PEA. Methods: A prospective single center observational study was conducted from May to November 2023 in patients that underwent PEA. Hemodynamic and biological data were collected from arterial and venous blood gas after the induction of general anesthesia and after CPB weaning. Results: Forty-nine patients were included in the final analysis. The mean age was 57 (±14.3) years, and 30 (61%) patients were male. There was no significant change in VO2 and DO2 (O2 delivery) after CPB weaning (VO2=104.5 (±45.9) vs 110.5 (±30.4) ml of O2/min/m2; p=0.33; DO2=426.1 (±166.3) vs 398.1 (±109.4) ml of O2/min/m2; p=0.18 respectively). There was a weak correlation between CPB duration and VO2 following CPB weaning (R=0.41; p=0.008). No correlation between the duration of aortic cross clamp time, the duration of circulatory arrest, and post CPB VO2 were found (R=0.22; p=0.14 and R=0.22; p=0.10, respectively). Conclusion: There was no significant increase in VO2 and DO2 after deep hypothermic circulatory arrest PEA surgery.

LI Baixun, LIU Tianxu, HUANG Liqin et al.

Hepatitis E is an acute and self-limiting viral hepatitis caused by the hepatitis E virus (HEV). It has a higher mortality rate among immunosuppressed patients and pregnant women infected with HEV. Although HEV infections in humans are mostly caused by contaminated water or food worldwide, the incidence of transfusion-transmitted hepatitis E is continuously rising. Additionally, the prevalence of serum anti-HEV IgG in the blood donors in China is at a relatively high level, making it worth considering screening blood donors for HEV. This article briefly reviews the globally reported cases of transfusion-transmitted hepatitis E and the HEV screening strategies for blood donations.

Ayyüce Begüm Bektaş, Mithat Gönen

This paper claims that machine learning models deployed in high stakes domains such as medicine must be interpretable, shareable, reproducible and accountable. We argue that these principles should form the foundational design criteria for machine learning algorithms dealing with critical medical data, including survival analysis and risk prediction tasks. Black box models, while often highly accurate, struggle to gain trust and regulatory approval in health care due to a lack of transparency. We discuss how intrinsically interpretable modeling approaches (such as kernel methods with sparsity, prototype-based learning, and deep kernel models) can serve as powerful alternatives to opaque deep networks, providing insight into biomedical predictions. We then examine accountability in model development, calling for rigorous evaluation, fairness, and uncertainty quantification to ensure models reliably support clinical decisions. Finally, we explore how generative AI and collaborative learning paradigms (such as federated learning and diffusion-based data synthesis) enable reproducible research and cross-institutional integration of heterogeneous biomedical data without compromising privacy, hence shareability. By rethinking machine learning foundations along these axes, we can develop medical AI that is not only accurate but also transparent, trustworthy, and translatable to real-world clinical settings.

Manuel Baltieri, Martin Biehl, Matteo Capucci et al.

The internal model principle, originally proposed in the theory of control of linear systems, nowadays represents a more general class of results in control theory and cybernetics. The central claim of these results is that, under suitable assumptions, if a system (a controller) can regulate against a class of external inputs (from the environment), it is because the system contains a model of the system causing these inputs, which can be used to generate signals counteracting them. Similar claims on the role of internal models appear also in cognitive science, especially in modern Bayesian treatments of cognitive agents, often suggesting that a system (a human subject, or some other agent) models its environment to adapt against disturbances and perform goal-directed behaviour. It is however unclear whether the Bayesian internal models discussed in cognitive science bear any formal relation to the internal models invoked in standard treatments of control theory. Here, we first review the internal model principle and present a precise formulation of it using concepts inspired by categorical systems theory. This leads to a formal definition of ``model'' generalising its use in the internal model principle. Although this notion of model is not a priori related to the notion of Bayesian reasoning, we show that it can be seen as a special case of possibilistic Bayesian filtering. This result is based on a recent line of work formalising, using Markov categories, a notion of ``interpretation'', describing when a system can be interpreted as performing Bayesian filtering on an outside world in a consistent way.

Zhihui Xie, Jizhou Guo, Tong Yu et al.

Large language models (LLMs) have demonstrated impressive capabilities in various reasoning tasks, aided by techniques like chain-of-thought prompting that elicits verbalized reasoning. However, LLMs often generate text with obvious mistakes and contradictions, raising doubts about their ability to robustly process and utilize generated rationales. In this work, we investigate reasoning in LLMs through the lens of internal representations, focusing on how these representations are influenced by generated rationales. Our preliminary analysis reveals that while generated rationales improve answer accuracy, inconsistencies emerge between the model's internal representations in middle layers and those in final layers, potentially undermining the reliability of their reasoning processes. To address this, we propose internal consistency as a measure of the model's confidence by examining the agreement of latent predictions decoded from intermediate layers. Extensive empirical studies across different models and datasets demonstrate that internal consistency effectively distinguishes between correct and incorrect reasoning paths. Motivated by this, we propose a new approach to calibrate reasoning by up-weighting reasoning paths with high internal consistency, resulting in a significant boost in reasoning performance. Further analysis uncovers distinct patterns in attention and feed-forward modules across layers, providing insights into the emergence of internal inconsistency. In summary, our results demonstrate the potential of using internal representations for self-evaluation of LLMs. Our code is available at github.com/zhxieml/internal-consistency.

Sydney Anuyah, Mallika K Singh, Hope Nyavor

The integration of artificial intelligence [AI] into clinical trials has revolutionized the process of drug development and personalized medicine. Among these advancements, deep learning and predictive modelling have emerged as transformative tools for optimizing clinical trial design, patient recruitment, and real-time monitoring. This study explores the application of deep learning techniques, such as convolutional neural networks [CNNs] and transformerbased models, to stratify patients, forecast adverse events, and personalize treatment plans. Furthermore, predictive modelling approaches, including survival analysis and time-series forecasting, are employed to predict trial outcomes, enhancing efficiency and reducing trial failure rates. To address challenges in analysing unstructured clinical data, such as patient notes and trial protocols, natural language processing [NLP] techniques are utilized for extracting actionable insights. A custom dataset comprising structured patient demographics, genomic data, and unstructured text is curated for training and validating these models. Key metrics, including precision, recall, and F1 scores, are used to evaluate model performance, while trade-offs between accuracy and computational efficiency are examined to identify the optimal model for clinical deployment. This research underscores the potential of AI-driven methods to streamline clinical trial workflows, improve patient-centric outcomes, and reduce costs associated with trial inefficiencies. The findings provide a robust framework for integrating predictive analytics into precision medicine, paving the way for more adaptive and efficient clinical trials. By bridging the gap between technological innovation and real-world applications, this study contributes to advancing the role of AI in healthcare, particularly in fostering personalized care and improving overall trial success rates.

Peng Xu, Hongjin Wu, Jinle Wang et al.

This paper details a technical plan for building a clinical case database for Traditional Chinese Medicine (TCM) using web scraping. Leveraging multiple platforms, including 360doc, we gathered over 5,000 TCM clinical cases, performed data cleaning, and structured the dataset with crucial fields such as patient details, pathogenesis, syndromes, and annotations. Using the $Baidu\_ERNIE\_Speed\_128K$ API, we removed redundant information and generated the final answers through the $DeepSeekv2$ API, outputting results in standard JSON format. We optimized data recall with RAG and rerank techniques during retrieval and developed a hybrid matching scheme. By combining two-stage retrieval method with keyword matching via Jieba, we significantly enhanced the accuracy of model outputs.

Pedro Zuidberg Dos Martires, Vincent Derkinderen, Luc De Raedt et al.

Recent developments in AI have reinvigorated pursuits to advance the (life) sciences using AI techniques, thereby creating a renewed opportunity to bridge different fields and find synergies. Headlines for AI and the life sciences have been dominated by data-driven techniques, for instance, to solve protein folding with next to no expert knowledge. In contrast to this, we argue for the necessity of a formal representation of expert knowledge - either to develop explicit scientific theories or to compensate for the lack of data. Specifically, we argue that the fields of knowledge representation (KR) and systems biology (SysBio) exhibit important overlaps that have been largely ignored so far. This, in turn, means that relevant scientific questions are ready to be answered using the right domain knowledge (SysBio), encoded in the right way (SysBio/KR), and by combining it with modern automated reasoning tools (KR). Hence, the formal representation of domain knowledge is a natural meeting place for SysBio and KR. On the one hand, we argue that such an interdisciplinary approach will advance the field SysBio by exposing it to industrial-grade reasoning tools and thereby allowing novel scientific questions to be tackled. On the other hand, we see ample opportunities to move the state-of-the-art in KR by tailoring KR methods to the field of SysBio, which comes with challenging problem characteristics, e.g. scale, partial knowledge, noise, or sub-symbolic data. We stipulate that this proposed interdisciplinary research is necessary to attain a prominent long-term goal in the health sciences: precision medicine.

Mustafa Ozmen , Cesur Gumus, Eda Otman et al.

Sandeep Padala, Jorge Cortes

Asciminib is a novel tyrosine kinase inhibitor (TKI) that specifically targets the myristoyl pocket. It has increased selectivity and potent activity against BCR-ABL1 and the mutants that most frequently prevent the activity of the ATPbinding competitive inhibitors. Results for clinical trials in patients with chronic myeloid leukemia that have received two or more TKI (randomized against bosutinib) or who have a T315I mutation (single arm study) have shown high levels of activity and a favorable toxicity profile. Its approval has offered new options for patients with these disease features. There are, however, a number of unanswered questions that remain to be defined, including the optimal dose, understanding the mechanisms of resistance, and, importantly, how it compares to ponatinib in these patient populations for whom we now have these two options available. Ultimately, a randomized trial is needed to answer questions to which we currently offer speculative informed guesses. The novelty of its mechanism of action and the exciting early data offer the potential for asciminib to address some of the remaining needs in the management of patients with chronic myeloid leukemia, including second-line therapy after resistance to a front-line second-generation TKI and improving successful treatment-free remission. Multiple studies are ongoing in these areas, and one can only hope that the desired randomized trial comparing asciminib to ponatinib will be conducted soon.

Corinne E. Praska, Riccardo Tamburrini, Juan Sebastian Danobeitia et al.

Organ transplantation is characterized by a sequence of steps that involve operative trauma, organ preservation, and ischemia-reperfusion injury in the transplant recipient. During this process, the release of damage-associated molecular patterns (DAMPs) promotes the activation of innate immune cells via engagement of the toll-like receptor (TLR) system, the complement system, and coagulation cascade. Different classes of effector responses are then carried out by specialized populations of macrophages, dendritic cells, and T and B lymphocytes; these play a central role in the orchestration and regulation of the inflammatory response and modulation of the ensuing adaptive immune response to transplant allografts. Organ function and rejection of human allografts have traditionally been studied through the lens of adaptive immunity; however, an increasing body of work has provided a more comprehensive picture of the pivotal role of innate regulation of adaptive immune responses in transplant and the potential therapeutic implications. Herein we review literature that examines the repercussions of inflammatory injury to transplantable organs. We highlight novel concepts in the pathophysiology and mechanisms involved in innate control of adaptive immunity and rejection. Furthermore, we discuss existing evidence on novel therapies aimed at innate immunomodulation and how this could be harnessed in the transplant setting.

Lyne Moser, Maru Sarazola, Paula Verdugo

Given a cartesian closed category $\mathcal{V}$, we introduce an internal category of elements $\int_\mathcal{C} F$ associated to a $\mathcal{V}$-functor $F\colon \mathcal{C}^{\mathrm{op}}\to \mathcal{V}$. When $\mathcal{V}$ is extensive, we show that this internal Grothendieck construction gives an equivalence of categories between $\mathcal{V}$-functors $\mathcal{C}^{\mathrm{op}}\to \mathcal{V}$ and internal discrete fibrations over $\mathcal{C}$, which can be promoted to an equivalence of $\mathcal{V}$-categories. Using this construction, we prove a representation theorem for $\mathcal{V}$-categories, stating that a $\mathcal{V}$-functor $F\colon \mathcal{C}^{\mathrm{op}}\to \mathcal{V}$ is $\mathcal{V}$-representable if and only if its internal category of elements $\int_\mathcal{C} F$ has an internal terminal object. We further obtain a characterization formulated completely in terms of $\mathcal{V}$-categories using shifted $\mathcal{V}$-categories of elements. Moreover, in the presence of $\mathcal{V}$-tensors, we show that it is enough to consider $\mathcal{V}$-terminal objects in the underlying $\mathcal{V}$-category $\mathrm{Und}\int_\mathcal{C} F$ to test the representability of a $\mathcal{V}$-functor $F$. We apply these results to the study of weighted $\mathcal{V}$-limits, and also obtain a novel result describing weighted $\mathcal{V}$-limits as certain conical internal limits.

Chris Solomou

The introduction of Large Language Models (LLMs), and the vast volume of publicly available medical data, amplified the application of NLP to the medical domain. However, LLMs are pretrained on data that are not explicitly relevant to the domain that are applied to and are often biased towards the original data they were pretrained upon. Even when pretrained on domainspecific data, these models typically require time-consuming fine-tuning to achieve good performance for a specific task. To address these limitations, we propose an alternative approach that achieves superior performance while being computationally efficient. Specifically, we utilize keywords to train a deep learning architecture that outperforms a language model pretrained on a large corpus of text. Our proposal does not require pretraining nor fine-tuning and can be applied directly to a specific setting for performing multi-label classification. Our objective is to automatically assign a new patient to the specialty of the medical professional they require, using a dataset that contains medical transcriptions and relevant keywords. To this end, we fine-tune the PubMedBERT model on this dataset, which serves as the baseline for our experiments. We then twice train/fine-tune a DNN and the RoBERTa language model, using both the keywords and the full transcriptions as input. We compare the performance of these approaches using relevant metrics. Our results demonstrate that utilizing keywords for text classification significantly improves classification performance, for both a basic DL architecture and a large language model. Our approach represents a promising and efficient alternative to traditional methods for finetuning language models on domain-specific data and has potential applications in various medical domains

Belén Díaz‐Antón, Rodrigo Madurga, Blanca Zorita et al.

Abstract Aims To evaluate echocardiographic and biomarker changes during chemotherapy, assess their ability to early detect and predict cardiotoxicity and to define the best time for their evaluation. Methods and results Seventy‐two women with breast cancer (52 ± 9.8 years) treated with anthracyclines (26 also with trastuzumab), were evaluated for 14 months (6 echocardiograms/12 laboratory tests). We analysed: high‐sensitivity cardiac troponin T, NT‐proBNP, global longitudinal strain (GLS), left ventricle end‐systolic volume (LVESV), left ventricle end‐diastolic volume (LVEDV), and left ventricular ejection fraction (LVEF). Cardiotoxicity was defined as a reduction in LVEF>10% compared with baseline with LVEF<53%. High‐sensitivity troponin T levels rose gradually reaching a maximum peak at 96 ± 13 days after starting chemotherapy (P < 0.001) and 62.5% of patients presented increased values during treatment. NT‐proBNP augmented after each anthracycline cycle (mean pre‐cycle levels of 72 ± 68 pg/mL and post‐cycle levels of 260 ± 187 pg/mL; P < 0.0001). Cardiotoxicity was detected in 9.7% of patients (mean onset at 5.2 months). In the group with cardiotoxicity, the LVESV was higher compared with those without cardiotoxicity (40 mL vs. 29.5 mL; P = 0.045) at 1 month post‐anthracycline treatment and the decline in GLS was more pronounced (−17.6% vs. −21.4%; P = 0.03). Trastuzumab did not alter serum biomarkers, but it was associated with an increase in LVESV and LVEDV (P < 0.05). While baseline LVEF was an independent predictor of later cardiotoxicity (P = 0.039), LVESV and GLS resulted to be early detectors of cardiotoxicity [odds ratio = 1.12 (1.02–1.24), odds ratio = 0.66 (0.44–0.92), P < 0.05] at 1 month post‐anthracycline treatment. Neither high‐sensitivity troponin T nor NT‐proBNP was capable of predicting subsequent cardiotoxicity. Conclusions One month after completion of anthracycline treatment is the optimal time to detect cardiotoxicity by means of imaging parameters (LVESV and GSL) and to determine maximal troponin rise. Baseline LVEF was a predictor of later cardiotoxicity. Trastuzumab therapy does not affect troponin values hence imaging techniques are recommended to detect trastuzumab‐induced cardiotoxicity.

Ahmad Chaddad, Qizong lu, Jiali Li et al.

Artificial intelligence (AI) continues to transform data analysis in many domains. Progress in each domain is driven by a growing body of annotated data, increased computational resources, and technological innovations. In medicine, the sensitivity of the data, the complexity of the tasks, the potentially high stakes, and a requirement of accountability give rise to a particular set of challenges. In this review, we focus on three key methodological approaches that address some of the particular challenges in AI-driven medical decision making. (1) Explainable AI aims to produce a human-interpretable justification for each output. Such models increase confidence if the results appear plausible and match the clinicians expectations. However, the absence of a plausible explanation does not imply an inaccurate model. Especially in highly non-linear, complex models that are tuned to maximize accuracy, such interpretable representations only reflect a small portion of the justification. (2) Domain adaptation and transfer learning enable AI models to be trained and applied across multiple domains. For example, a classification task based on images acquired on different acquisition hardware. (3) Federated learning enables learning large-scale models without exposing sensitive personal health information. Unlike centralized AI learning, where the centralized learning machine has access to the entire training data, the federated learning process iteratively updates models across multiple sites by exchanging only parameter updates, not personal health data. This narrative review covers the basic concepts, highlights relevant corner-stone and state-of-the-art research in the field, and discusses perspectives.

Bojana Femić, Enrico Ghiorzi

We introduce categories $\M$ and $§$ internal in the tricategory $\Bicat_3$ of bicategories, pseudofunctors, pseudonatural transformations and modifications, for matrices and spans in a 1-strict tricategory $V$. Their horizontal tricategories are the tricategories of matrices and spans in $V$. Both the internal and the enriched constructions are tricategorifications of the corresponding constructions in 1-categories. Following \cite{FGK} we introduce monads and their vertical morphisms in categories internal in tricategories. We prove an equivalent condition for when the internal categories for matrices $\M$ and spans $§$ in a 1-strict tricategory $V$ are equivalent, and deduce that in that case their corresponding categories of (strict) monads and vertical monad morphisms are equivalent, too. We prove that the latter categories are isomorphic to those of categories enriched and discretely internal in $V$, respectively. As a byproduct of our tricategorical constructions we recover some results from \cite{Fem}. Truncating to 1-categories we recover results from \cite{CFP} and \cite{Ehr} on the equivalence of enriched and discretely internal 1-categories.

Sanjay Dixit, Aung Lin

Héctor Rodríguez, Rafael Prados-Rosales, José Luis Lavín et al.

Luis Manuel Limas Solano, Carlos Ernesto Vargas Niño, Diana Carolina Valbuena Rodríguez et al.

La microbiota intestinal sana se define a partir de la presencia de grupos de microorganismos que potencian el metabolismo del huésped. Estos microorganismos le confieren resistencia ante las infecciones, así como ante procesos inflamatorios y frente al desarrollo de neoplasias o autoinmunidad. Además, favorecen las funciones endocrinas y colaboran con la función neurológica a través del eje intestino-cerebro. Por otro lado, el trasplante de microbiota fecal consiste en la introducción de una suspensión de materia fecal de un donante sano en el tracto gastrointestinal de otra persona, que generalmente es un paciente que presenta una patología concreta. Esto se realiza con el fin de manipular la composición de la microbiota del destinatario y contribuir al tratamiento de su problema. El concepto de trasplante de microbiota fecal rompe con la consideración tradicional de las bacterias como elementos dañinos y presta atención a las que, probablemente, son las más subvaloradas de las excretas del cuerpo humano: las heces. En efecto, se ha evidenciado su alta eficacia y el procedimiento es reconocido por el número de pacientes a los que ha ayudado, que se puede ya cifrar en miles. El objetivo de esta revisión de literatura fue describir aspectos básicos para comprender el trasplante de microbiota fecal enfocado al tratamiento de infecciones producidas por Clostridioides difficile.

Xin Yin, Jinghua Xia, Ying Sun et al.

Abstract Background CHCHD2 was identified a novel cell migration-promoting gene, which could promote cell migration and altered cell adhesion when ectopically overexpressed in NIH3T3 fibroblasts, and it was identified as a protein necessary for OxPhos function as well. However, the clinic relevance of CHCHD2 expression in NSCLC remains unclear. Here we assumed that CHCHD2 expression would accompanies the expression of HIF-1α to response hypoxia in the occurrence of NSCLC. Methods In order to verify this hypothesis, correlations among the expression levels of CHCHD2 and HIF-1α were detected and analyzed in 209 pair cases of NSCLC. The expression and location of these molecules were assessed using Immunohistochemistry, immunohistofluorescence, qRT-PCR and western blotting. The differences and correlations of the expression of these two molecules with clinical pathological characteristics in NSCLC were statistically analyzed using Wilcoxon (W) text, Mann-Whitney U, Kruskal-Wallis H and cross-table tests. Kaplan-Meier survival analysis and Cox proportional hazards models were used to estimate the effect of the expression of CHCHD2 and HIF-1α on the patients’ survival. Results Data showed that CHCHD2 and HIF-1α expression were higher in NSCLC than in normal tissues (all P = 0.000). CHCHD2 expression was significantly related with smoking, tumor size, differentiation degree, TNM Stage, lymph metastasis (all P<0.05). The HIF-1α expression was significantly associated with smoking, tumor category, differentiation degree, TNM Stage, Lymph metastasis (all P<0.05). There was a marked correlation of CHCHD2 and HIF-1α expression with histological type, differentiation and lymph metastasis of NSCLC (all P<0.05, rs >0.3). Immunohistofluorescence showed that there were co-localization phenomenon in cytoplasm and nucleus between CHCHD2 and HIF-1α expression. NSCLC patients with higher CHCHD2 and HIF-1α expression had a significantly worse prognosis than those with lower CHCHD2 and HIF-1α expression (all P = 0.0001; log-rank test). The multivariate analysis indicated that CHCHD2 expression was an independent prognostic factor in NSCLC (hazard ratio [HR], 0.492, P = 0.001). Conclusion Our results indicate that over-expression of CHCHD2 would promote the expression of HIF-1α to adapt the hypoxia microenviroment in NSCLC and CHCHD2 could serves as a prognostic biomarker in NSCLC.